Baker Gregory J, Novikov Edward, Coy Shannon, Chen Yu-An, Hug Clemens B, Ahmed Zergham, Cajas Ordóñez Sebastián A, Huang Siyu, Yapp Clarence, Sokolov Artem, Pfister Hanspeter, Santagata Sandro, Sorger Peter K
Laboratory of Systems Pharmacology, Harvard Medical School, Boston, MA.
Ludwig Center for Cancer Research at Harvard, Harvard Medical School, Boston, MA.
bioRxiv. 2025 Jun 27:2025.06.23.661064. doi: 10.1101/2025.06.23.661064.
Spatial proteomics (highly multiplexed tissue imaging) provides unprecedented insight into the types, states, and spatial organization of cells within preserved tissue environments. To enable single-cell analysis, high-plex images are typically segmented using algorithms that assign marker signals to individual cells. However, conventional segmentation is often imprecise and susceptible to signal spillover between adjacent cells, interfering with accurate cell type identification. Segmentation-based methods also fail to capture the morphological detail that histopathologists rely on for disease diagnosis and staging. Here, we present a method that combines unsupervised, pixel-level machine learning using autoencoders with traditional segmentation to generate single-cell data that captures information on protein abundance, morphology, and local neighborhood in a manner analogous to human experts while overcoming the problem of signal spillover. The result is a more accurate and nuanced characterization of cell types and states than segmentation-based analysis alone.
空间蛋白质组学（高度多重组织成像）为保存的组织环境中细胞的类型、状态和空间组织提供了前所未有的见解。为了实现单细胞分析，高多重图像通常使用将标记信号分配给单个细胞的算法进行分割。然而，传统的分割往往不准确，并且容易受到相邻细胞之间信号溢出的影响，干扰准确的细胞类型识别。基于分割的方法也无法捕捉组织病理学家用于疾病诊断和分期的形态学细节。在这里，我们提出了一种方法，该方法将使用自动编码器的无监督像素级机器学习与传统分割相结合，以生成单细胞数据，该数据以类似于人类专家的方式捕获有关蛋白质丰度、形态和局部邻域的信息，同时克服信号溢出问题。与仅基于分割的分析相比，结果是对细胞类型和状态进行更准确、更细致入微的表征。
